Transforming growth factor-bs (TGF-bs) are potent inhibitors of epithelial cell growth. Recently components of the TGF-b response path have been shown to be diminished or absent in a number of human malignancies, implicating loss of TGF-b function as one mechanism contributing to tumor development. However, TGF-b expression is often upregulated in advanced human cancers suggesting that the role played by the TGF-b system may be complex. We and others have proposed that TGF-bs function as tumor suppressors early in the tumorigenic process and as pro-metastatic agents in late-stage disease. Since the effects of TGF-b are highly context-dependent, where possible we have chosen to study this hypothesis in vivo in the whole animal, so that all the complex contextual cues are maintained. Our approach has been to generate and analyze genetically engineered mice in which TGF-b function is experimentally compromised. In a complementary approach, we genetically modify TGF-b function in cell lines and determine the effect on tumorigenicity by xenografting these into nude mice. We focus particularly on breast cancer. Using a series of human breast-derived cell lines representing different stages of the tumorigenic process we have set up a model system in which we have demonstrated that TGF-bs do indeed switch from tumor suppressor to oncogene in the late stages of breast cancer, thus confirming our original hypothesis. We are exploiting this model system to analyze the underlying biological and molecular mechanisms. To address whether neutralization of TGF-b might be a useful therapeutic approach for advanced cancer, we have generated transgenic mice expressing a soluble TGF-b antagonist. We have shown that these mice are protected against metastasis, both in metastatic breast cancer and metastatic melanoma model systems. Unexpectedly, there were essentially no adverse side effects, so this form of TGF-b antagonist shows promise for clinical use in the prevention of metastasis. Finally, we are working with mice in which components of the TGF-b signal transduction pathway have been genetically knocked out. This will allow us to determine whether the tumor suppressor and oncogenic activities of TGF-b are mediated by distinct signaling pathways. Results from all these experiments should give clinically useful insights into the functions of TGF-bs during tumor initiation, promotion and progression, and illuminate how the system could be most effectively manipulated in novel chemopreventive or therapeutic strategies.